The Center for Autonomic Computing (CAC), an NSF Research Center funded by the I/UCRC pro-

gram, combines resources from universities, private companies, and the federal government to

conduct fundamental research on making all kinds of computer systems and applications - from

humble desktop computers to air traffic control systems - more reliable, more secure, and more

efficient.

Autonomic computing (AC) denotes a broad area of scientific and engineering research on meth-

ods, architectures and technologies for the design, implementation, integration and evaluation of

special and general-purpose computing systems, components and applications that are capable of

autonomously achieving desired behaviors. AC systems aim to be self-managed in order to enable

independent operation, minimize cost and risk, accommodate complexity and uncertainty and

enable systems of systems with large numbers of components. Hence, system integration and

automation of management are important areas of research whose contexts subsume other AC

research topics. These might include, to varying degrees, self-organization, self-healing, self-

optimization (e.g. for power or speed), self-protection and other similar behaviors. CAC research

activities will advance several disciplines that impact the specification, design, engineering, and

integration of autonomic computing and information processing systems. They include design and

evaluation methods, algorithms, architectures, information processing, software, mathematical

foundations and benchmarks for autonomic systems. Solutions will be studied at different levels

of both centralized and distributed systems, including the hardware, networks, storage, middle-

ware, services and information layers. Collectively, the participating universities have research

and education programs whose strengths cover the technical areas of the center. Within this broad

scope, the specific research activities will vary over time as a reflection of center member needs

and the evolution of the field of autonomic computing.

Inside this issue:

Introduction 1

From the Director’s desk

2

Ongoing projects— I

3

Ongoing projects— II

4

Ongoing projects— III

5

Ongoing projects— IV

6

Key Research Areas & SC’08

7

Membership 8

Sponsored by the

National Science

Foundation Office of

Industrial Innovat ion

and Partnerships

Industry/University

Cooperative Research

Center (I/UCRC)

Program under award

0758596.

CoRE TeCH F a l l 2 0 0 8 V o l u m e 1 , I s s u e 1

Address:

NSF Center for

Autonomic Computing,

Rutgers University,

94 Brett Road ,

Piscataway ,

NJ 08854-8058

The NSF Center for Autonomic Computing at Rutgers

The Rutgers Green Computing Initiative

(nsfcac.rutgers.edu/greencomputing/)

Fall 2008 CAC Meeting, University of Arizona

(nsfcac.arizona.edu/nov08-meeting.html)

IBM Faculty Award awarded to Professor

Manish Parashar, Director, NSF CAC, Rutgers

for research on autonomic data extraction,

streaming and in-transit data manipulation.

Two day hands-on workshop on ‘Programming

the IBM CELL BE processor’ on Sept 18 & 19,’08

CAC @Rutgers part of multiple prestigious

grants from NSF and DOE:

"Actively Managing Data Movement with Models - Taming High Performance Data Com-

munications in Exascale Machines," NSF HECURA, 08/08 - 07/11.

"Computational Models for Evaluating Long Term CO2 Storage in Saline Aquifers," NSF

CDI, 09/01/08 - 08/31/12 .

" Cross-layer Research on management of Virtualized Datacenters," NSF I/UCRC Supple-

ment, 08/08 - 12/09.

http://nsfcac.rutgers.edu/

Partner sites:

University of Florida

(www.nsfcac.org)

University of Arizona

(nsfcac.arizona.edu)

What’s new ?

P a g e 2

From the Director’s desk It is my pleasure to introduce to you the National Science Foundation Center for Autonomic

Computing (NSF CAC) - a center funded through the NSF Industry/University Cooperative Research

Centers program, industry, government agencies and matching funds from member universities,

which currently include the University of Florida (Lead), the University of Arizona and Rutgers, The

State University of New Jersey. The mission of the center is to advance the knowledge of designing

Information Technology (IT) systems and services to make them self-governed and self-managed.

This will be achieved through developing innovative designs, programming paradigms and capabili-

ties for computing systems.

The explosive growth of IT infrastructures, coupled with the diversity of their components

and a shortage of skilled IT workers, have resulted in systems whose control and timely management

exceeds human ability. Current IT management solutions are costly, ineffective and labor intensive.

According to estimates by the International Data Corporation (IDC), worldwide costs of server man-

agement and administration will reach approximately US $150 Billion by 2010 and are estimated to

represent more than 60% of the total spending needed to establish, maintain and operate large IT

infrastructures. Autonomic Computing (AC) models IT infrastructures and their applications as

closed-loop control systems that need to be continuously monitored and analyzed, and subject to cor-

rective actions whenever any of the desired behavior properties (e.g., performance, reliability, secu-

rity) are violated. Such AC techniques are inspired by strategies used by biological systems to deal

with complexity, dynamism, heterogeneity and uncertainty. The design space for designing AC sys-

tems spans multiple disciplines such as distributed computing, virtualization, control theory, artificial

intelligence, statistics, software architectures, mathematical programming, and networking. Our re-

search efforts will accelerate the research and development of core autonomic technologies and ser-

vices. It will also establish strong collaboration programs with the US IT industry, and specifically

New Jersey based industries, to develop next generation information and communication technolo-

gies and services that are inherently autonomic.

In this newsletter, we highlight our ongoing autonomic research activities and how to apply

them to wide range of applications/domains with profound impact on economy and industry. These

include autonomic IT infrastructures (specifically, autonomic computing engines that support auto-

nomic cloudbursts and system-level acceleration), autonomic management of IT infrastructure

(specifically, performance/power of large-scale datacenters, robust clustering analysis and online self

-monitoring, autonomic management and dynamic data-driven management of instrumented data-

centers), data communication and streaming (autonomic data transport and in-transit manipulation

and high-throughput asynchronous data transfers), and applications of these technologies to a range

of domains including computational finance (online risk analytics), medical informatics (cancer TMA

analysis, REMD for drug design), and science/engineering (instrumented oilfield management, trans-

portation, ocean observatories).

In summary, the NSF-CAC will not only advance the science of autonomic computing, but

also accelerate the transfer of technology to industry and contribute to the education of a workforce

capable of designing and deploying autonomic computing systems to many sectors---from homeland

security and defense, to business agility to the science of global change. Furthermore, the center will

leave its mark on the education and training of a new generation of scientists that have the skills and

know-how to create knowledge in new transformative ways.

As we move forward, we would like to invite you to join the center so together we can de-

velop innovative autonomic technologies and services that will revolutionize how to design and de-

ploy next generation computation, information and communications services.

by

Manish Parashar,

Director,

NSFCAC, Rutgers

C o R E T e C H F a l l 2 0 0 8

― The mission of

the center is to

advance the

knowledge of

designing

Information

Technology (IT)

systems and

services to make

them self-

governed and

self-managed. ‖

P a g e 3

Selected Ongoing Projects

A self-monitoring system is able to observe and analyze system state and behavior, to dis-cover anomalies or violations, and to notify autonomic or human administrators in a timely manner so that appropriate management actions can be effectively applied. Implementing the analysis of sys-tem state in a decentralized and in-network fashion (using network resources and minimal extrane-ous information) ensures computational tractability and acceptable response times, enabling automated and online system management. However, because self-monitoring mechanisms are sub-ject to the same failures that occur in the network that they are helping to manage, the robustness of these mechanisms is of great importance to ensure overall system reliability.

The main contribution of this work is the formulation and validation of a robust decentral-ized data analysis mechanism that applies density-based clustering techniques to identify anomalies and clusters of arbitrary size and shape in monitoring data. Clustering data is given in the form of periodic behavior and operational status update events from system components, defined in terms of known attributes. The event attributes are used to construct a multidimensional coordinate space, which is then used to measure the similarity of events. Components that behave in a similar fashion can then be identified by the clusters formed by their status events in this space, while devices with abnormal behavior will produce isolated events. The clustering algorithm requires minimal computa-tion at processing nodes, which makes it suitable for online execution.

The robustness of the decentralized mechanisms is dealt with at three levels. First, we as-sume that the connectivity of the network is maintained despite node failures through self-healing mechanisms provided at the overlay level. Next, at the data messaging level, we use replication to prevent the loss of the events required for the clustering analysis. To minimize the overhead incurred by replication, data is selectively replicated at nodes based on their probability of failure, which is obtained by maintaining a failure history and calculated using an appropriate failure model. The se-lectivity of replication can be further aided by information available at the analysis level. Because the primary focus of the clustering analysis is on anomaly detection, only points that are most likely to be anomalies should be replicated. This can be predicted given previous clusters and anomalies observed in the system.

This work is part of an ongoing effort to create tools for integrated data analysis for the auto-nomic management of performance, security/trust and reliability at a system level. Current and fu-ture efforts include improving cluster descriptions produced by the algorithm for effective profiling of system behavior and developing predictive system models of distributed system state. We plan to combine these mechanisms with tools for defining and conditioning the application of system policies with these profiles and state predictions for autonomic resource management and provisioning, usage control and monitoring, and trust management and authentication.

by

Andres Quiroz

Robust Clustering Analysis for Self-Monitoring Distributed Systems

Principal

Researchers:

Andres Quiroz

Manish Parashar

Current

Collaborators:

Naveen Sharma

Nathan Gnana-

sambandam

References:

―Robust Cluster-

ing Analysis for

the Management

of Self-Monitoring

Distributed Sys-

tems,‖ A. Quiroz,

N. Gnanasamban-

dam, M. Parashar,

and N. Sharma.

To appear in

Journal of Cluster

Computing,

Springer 2008,

DOI: 10.1007/

s10586-008-0068

-5.

C o R E T e C H F a l l 2 0 0 8

Cluster Analysis for Anomaly Detection

* Define & divide an information space

* Analyze density of events in each region

* Recognize anomalies by regions with low density

* Charaterize clusters to obtain behavior profile

Selective replication for robustness

* Fault tolerance & replication mechanisms at

different levels

* Use of clustering results to reduce the overhead of

replication while maintaining accuracy

P a g e 4

Autonomic Management of Instrumented Datacenters

by

Nanyan Jiang

Principal

Researchers:

Nanyan Jiang

Manish Parashar

Dario Pompili

Technical advances are leading to a pervasive computational ecosystem that integrates com-puting infrastructures with embedded sensors and actuators, and giving rise to a new paradigm for monitoring, understanding, and managing natural and engineered systems – one that is information/data-driven and autonomic. The overarching goal of this research is to develop sensor system middle-ware and programming support that will enable distributed networks of sensors to function, not only as passive measurement devices, but as intelligent data processing instruments, capable of data qual-ity assurance, statistical synthesis and hypotheses testing as they stream data from the physical envi-ronment to the computational world.

This research investigates a programming system to support the development of in-network data processing mechanisms, and enable applications to discover, query, in-teract with, and control instru-mented physical systems, such as autonomic datacenter management system using semantically meaning-ful abstractions. This includes ab-stractions and runtime mechanisms for integrating sensor systems with applications processes, as well as for in-network data processing such as aggregation, adaptive interpolation and assimilation.

The proposed system enables sensor-driven autonomic manage-ment of instrumented datacenters at two levels. First, it provides pro-gramming abstractions for integrat-ing sensor systems with computa-tional models for applications proc-esses and with other application components in an end-to-end experi-ment. Second, it provides program-ming abstractions and system soft-ware support for developing in-network data processing mecha-nisms. Specifically for the latter, we explore the temporal and spatial cor-relation of sensor measurements in the targeted application domains to tradeoff between the complexity of coordination among sensor clusters and the savings that result from having fewer sensors for in-network processing, while main-taining an acceptable error threshold. Experimental results show that the proposed in-network mechanisms can facilitate the efficient usage of constraint resources and satisfy data requirement in the presence of dynamics and uncertainty.

In such a scenario, the programming system is used to enable the dynamic data-driven man-agement of an instrumented data center system, to optimize power consumption, efficiency of cooling system and job throughput. Sensor networks monitor temperature, humidity, and airflow in real time, provide non-intrusive and fine-grained data collection, and enable real-time processing. And these sensors are integrated with computational processes and job schedulers to take phenomenon, such as heat distribution and air flows into consideration, and to optimize data center performance in terms of energy consumption and throughput. More specifically, the scheduler uses real time infor-mation about data center operational conditions (e.g., temperature, humidity) from the sensors to generate optimal management policies. Furthermore, to enable timely response to environmental changes, in-network analysis uses localized temperature distribution to decide when and where to migrate jobs at runtime. Experimental results show that the provided programming system reduces overheads while achieving near optimal and timely management and control.

A Sensor-driven Data Center Management Scenario

C o R E T e C H F a l l 2 0 0 8

A Two-Level Autonomic Data Center

Management System

References:

N. Jiang and M. Parashar, ―Programming Support for Sen-sor-based Scien-tific Applica-tions,‖ Proceed-ings of the Next Generation Soft-ware (NGS) Workshop, (IPDPS 2008), Miami, FL, USA, IEEE Computer Society Press, April, 2008.

N. Jiang and M. Parashar, ―In-network Data E s t i m a t i o n Mechanisms for S e n s o r - d r i v e n Scientific Applica-tions,‖ Proceed-ings of the 15th IEEE Interna-tional Conference on High Perform-ance Computing (HiPC 2008) , Bangalore, India, December, 2008.

P a g e 5

Cluster-based data centers have become dominant computing platforms in industry and re-search for enabling complex and compute intensive applications. However, as scales, operating costs, and energy requirements increase, maximizing efficiency, cost-effectiveness, and utilization of these systems becomes paramount. Furthermore, the complexity, dynamism, and often time critical nature of application workloads makes on-demand scalability, integration of geographically distributed re-sources, and incorporation of utility computing services extremely critical. Finally, the heterogeneity and dynamics of the system, application, and computing environment require context-aware dynamic scheduling and runtime management.

A u to n omi c c lo u d bursts is the dynamic deploy-ment of a software application that runs on internal organiza-tional compute resources to a public cloud to address a spike in demand. Provisioning data center resources to handle sud-den and extreme spikes in de-mand is a critical requirement, and this can be achieved by combining both private data center resources and remote on-demand cloud resources such as Amazon EC2, which provides resizable computing capacity in the cloud.

This project envisions a computational engine that can enable autonomic cloud bursts capa-ble of: (1) Supporting dynamic utility-driven on-demand scale-out of resources and applications, where organizations incorporate computational resources based on perceived utility. These include resources within the enterprise and across virtual organizations, as well as from emerging utility com-puting clouds. (2) Enabling complex and highly dynamic application workflows consisting of hetero-geneous and coupled tasks/jobs through programming and runtime support for a range of computing patterns (e.g., master-slave, pipelined, data-parallel, asynchronous, system-level acceleration). (3) Integrated runtime management (including scheduling and dynamic adaptation) of the different di-mensions of application metrics and execution context. Context awareness includes system awareness to manage heterogeneous resource costs, capabilities, availabilities, and loads, application awareness to manage heterogeneous and dynamic application resources, data and interaction/coordination re-quirements, and ambient-awareness to manage the dynamics of the execution context.

As part of this project we have developed the Comet computing substrate, which provides a foundation and core capabilities for the envisioned autonomic computing engine. Comet supports different programming abstractions for parallel computing, including master/worker, BOT, work-flow, data parallel and asynchronous iterations, in a dynamic and widely distributed environment. It provides the abstraction of virtual semantic shared spaces that forms the basis for flexible scheduling, associative coordination, and content-based asynchronous and decoupled interactions. Comet builds on a self-organizing and fault-tolerant dynamic overlay of computing resources. It is currently de-ployed on a range of platforms, including local clusters, campus Grids, wide-area computing plat-forms (e.g., PlanetLab), Microsoft HPCS, and Amazon EC2, and supports a number of computational applications across multiple disciples including science & engineering, bio/medical-informatics and finance.

Ongoing efforts are focused on dynamically growing and shrinking clouds according to the workloads and optimization policies such as time (get the result as soon as possible with sufficient budget) or budget (get the result limiting costs to within a budget). Also we have working on robust-ness and security issues, e.g., isolating unsecured nodes in a cloud form sensitive data and application logic. Hence the autonomic cloud bursts can be possible in the aspect of secure reliable nodes and unsecured cloud nodes. Future works include utility-based and cooperative workflow scheduling models, scheduling and monitoring tasks, simultaneous failure management, and support for high-level application models such as Hadoop/MapReduce.

by

Hyunjoo Kim

Autonomic Cloud Bursts using Comet Autonomic Computing Engine

Principal

Researchers:

Hyunjoo Kim

Shivangi Chaud-

hari

Vamsi Kodama-

simham

Manish Parashar

Architectural Overview of Comet

C o R E T e C H F a l l 2 0 0 8

References:

Z. Li and M. Par ashar , " A C omputat i onal Infrastructure for Grid-based Asyn-chronous Parallel A p p li c a t i o n s, " Proceedings of the 16th Interna-tional Symposium o n H i g h -Performance Dis-tributed Comput-i ng ( HPDC ) , Monterey, CA, USA, pp. 229, June 2007.

P a g e 6

Emerging enterprise/Grid applications consist of complex workflows, which are composed of interacting components/services that are separated in space and time and execute in widely dis-tributed environments. Couplings and interactions between components/services in these applica-tions are varied, data intensive and time critical. As a result, high-through, low latency data acquisi-tion, data streaming and in-transit data manipulation is critical.

The goal of this project is to develop and deploy autonomic data management services that sup-port high throughput, low latency data streaming and in-transit data manipulation. Key attributes/requirements of the service include: (1) support for high-throughput, low-latency data transfers to enable near real-time access to the data, (2) ability to stream data over wide area network with shared resources and varying loads, and be able to maintain desired QoS, (3) minimal performance over-heads on the application, (4) adaptations to address dynamic application, system, and network states, (5) proactive control to prevent loss of data, and (6) effective management of in-transit processing while satisfying the above requirements.

The autonomic services ad-dress end-to-end QoS require-ments at two levels, which cooper-ate to address overall application constraints and QoS require-ments. The QoS management strategy at the application end-points combines model-based limited look-ahead controllers (LLC) and policy-based managers with adaptive multi-threaded buffer management. The applica-tion-level data streaming service consists of a service manager and an LLC controller.

The QoS manager monitors state and execution context, collects and reports runtime infor-mation, and enforces adaptation actions determined by its controller. In-transit data processing is achieved using a dynamic overlay of available resources in the data path between the source and the destination (e.g., workstations or small to medium clusters, etc.) with heterogeneous capabilities and loads. Note that these nodes may be shared across multiple applications flows. The goal of in-transit processing is to opportunistically process as much data as possible before the data reaches the sink, while ensuring that end-to-end timing constraints are satisfied. The combined constraints are cap-tured using a slack metric, which bounds the time available for data processing and transmission, such that the data reaches the sink in a timely manner. The in-transit nodes then use this slack metric to appropriately select in-transit resources from the dynamic overlay so as to maximize the data that is processed in-transit and consequently the quality of data reaching the destination.

Experiments with end-to-end cooperative data streaming demonstrated that adaptive proc-essing using the autonomic in-transit data processing service during congestions decreases the aver-age idle time per data block from 25% to 1%, thereby increasing utilization at critical times. Further-more, coupling end-point and in-transit management during congestion reduces average buffer occu-pancy at in-transit nodes from 80% to 60.8%, thereby reducing load and potential data loss, and in-creasing data quality at the destination. Ongoing work is focused on incorporating learning models for proactive management, and virtualization at in-transit nodes to improve utilization.

by

Ciprian Docan

Autonomic Data Streaming and In-Transit Processing

Principal

Researchers:

Viraj Bhat

Ciprian Docan

Manish Parashar

Current

Collaborators:

Scott Klasky

References:

―An Self-Managing Wide-Area Data Streaming Ser-vice,‖ V. Bhat, M. Parashar, H. Liu, M. Khandekar, N. Kandasamy, S. Klasky, and S. Abdelwahed, Cluster Comput-ing: The Journal of Networks, Soft-ware Tools, and Applications, Spe-cial Issue on Autonomic Com-puting, Kluwer, Volume 10, Issue 7, pp. 365 – 383 December 2007.

End-to-end Autonomic Streaming

C o R E T e C H F a l l 2 0 0 8

P a g e 7

C o R E T e C H F a l l 2 0 0 8

― SC is the premier international conference for high performance computing (HPC), networking, storage and analy-

sis. SC08 marked the 20th anniversary of the conference series, and featured the latest scientific and technical inno-

vations from around the world. The event had over 11,000 attendee and over 337 exhibitors. The CAC booth intro-

duced the center and displayed research results from the 3 sites. ‖

NSF CAC Booth at SC'08, Austin, TX

Computing Infrastructures

Autonomic Computing Engine and Autonomic Cloud Bursts

Autonomic Service Architecture for Self-Managing Applications

Programming Sensor-driven Autonomic Applications

Datacenter Management

Robust Clustering Analysis for Self-Monitoring Distributed Systems

Dynamic Data-Driven Management of Instrumented Datacenters

Cooperative Defense against Network Attacks

Data Communication/Streaming

Autonomic Data IO, Streaming and In-transit Processing

High-throughput Asynchronous Data Transfers

Applications

Computational Finance: Online Risk Analytics

Medical Informatics: Cancer TMA Analysis, REMD for Drug Design

Science/Engineering: Instrumented Oilfield, Transportation, Oceanography

Others

Hardware autonomics, sensing/actuation systems, etc.

Key Research Areas in Autonomic Computing

P a g e 8

CAC Personnel at Rutgers

Faculty

Dr. Manish Parashar

Dr. Dario Pompili

Dr. Michael Bushnell

Dr. David Foran

Dr. Hui Xiong

Post-docs

Hyunjoo Kim

System Administrator

Jim Housell

Graduate Students

Nanyan Jiang

Mingliang Wang

Andres Quiroz

Ciprian Docan

Shivangi Chaudhari

Richa Saraf

Siddharth Wagh

UG Students

Vamsi Kodamasimhan

Pramod Kulkarni

CAC members are afforded access to leading-edge

developments in autonomic computing and to knowl-

edge accumulated by academic researchers and other

industry partners. New members will join a growing

list of founding members that currently includes BAE

Systems, EWA Governemnt Systems, IBM, Intel,

Merrill-Lynch, Microsoft, Motorola, Northrop-

Grumman, NEC, Raytheon, Xerox, Avirtech, Citrix,

Imaginestics, and ISCA Technologies. Benefits of

membership include:

Collaboration with faculty, graduate students,

post-doctoral researchers and other center part-

ners

Choice of project topics to be funded by mem-

bers' own contributions

Formal periodic project reviews along with con-

tinuous informal interaction and timely access to

reports, papers and intellectual property gener-

ated by the center

Access to unique world-class equipment, facili-

ties, and other CAC infrastructure

Recruitment opportunities among excellent

graduate students

Leveraging of investments, projects and activities

by all CAC members

Spin-off initiatives leading to new partnerships,

customers or teaming for competitive proposals

Benefits of Membership

Funding

Founding members

Per NSF guidelines, industry and government contributions

in the form of annual CAC memberships ($35K/year per

regular membership), coupled with baseline funds from NSF

and university matching funds, directly support the Center's

expenses for personnel, equipment, travel, and supplies.

Annual Memberships provide funds to support the Center's

graduate students on a one-to-one basis, while NSF and

university funds support various other costs of operation.

Multiple annual memberships may be contributed by any

organization wishing to support multiple students and/or

projects. The initial operating budget for CAC is projected to

be approximately $1.5M/year, including NSF and universi-

ties contributions, in an academic environment that is very

cost effective. Thus, a single regular membership is an ex-

ceptional value. It represents less than 3% of the projected

annual budget of the Center yet reaps the full benefit of Cen-

ter activities, a research program that could be significantly

more expensive in an industry or government facility.

C o R E T e C H F a l l 2 0 0 8

To become a member, contact:

Director: Dr. Manish Parashar

Office: 628 CoRE, ECE Dept, Rutgers University Address: 94 Brett Road , Piscataway , NJ 08854-8058

Voice: (732) 445-5388 Email: parashar@rutgers.edu

Fax: (732) 445-0593 Website: http://nsfcac.rutgers.edu/